Foam Characteristics and Sensory Analysis of Arabica Coffee, Extracted by Espresso Capsule and Moka Methods

Round 1

Reviewer 1 Report

The article is interesting and broadens the knowledge about coffee and the methods of its preparation.

I have a few comments on some aspects of the study:

I am wondering about the research methodology. You probably want to show the influence of different coffee brewing pressures on cream and organoleptic properties.

First: why capsule espresso machines? I can guess the reasons, but it is worth writing about it clearly in the article. The popularity of these coffee machines is probably not a good, or at least the only reason. For example, I do not have a positive opinion of them. Please convince me that it is worth researching them.

Second: they are different devices. Is pressure the only factor that differentiates it? Isn't there the influence of other factors that differentiate the coffee from these machines? If so, the differences in coffee infusions from these devices cannot be explained by pressure alone. Your research methodology is defensible, but it needs to be written.

Third: why two types of coffee and why these? If you were to focus on one type of coffee, the pressure effects would be obvious and the results would be clear. Now these are the questions. Why only two species? Are they exhausting all the Arabica varieties? They are so different that the message of the differences is to be visible? And why exactly them? Because the manufacturer has them on offer? A very bad reason. In my opinion, the results shown in Figure 3 show well that this pressure is an ambiguous factor. And changing the type of coffee means that the results are not obvious. Describe it better in your conclusions. It begs for discussion. One species of Arabica is more influenced by pressure than others. Why? What is the conclusion? You have a little bit too few results for a full discussion, but you can signal that you see the problem. And you absolutely have to justify the reason why you took these two types of coffee and just these two. For what?

Fourth: You used a specific mineral water for the study. Did the mineralization of this water affect the results? Wouldn't a different mineral composition make your results different? Or maybe it has so few minerals that it has no significant effect on the properties of the infusion? It is worth writing.

In the study of sensory parameters, a ten-point scale of ratings assigned by panel participants was used. These ratings are not values ​​on a measurable scale! Rating is not a number. Has been given or assigned, not measured. It looks like a number, but it is not. It is a quality feature. They cannot be used to calculate measures and use tools appropriate to quantitative characteristics. Such as means, standard deviations or PCA. I know this is commonly done, but it is incorrect. Do you know of any studies that show that the values ​​of the scale you use are measurable? Or, it is not a mistake to use quantitative tools for these assessments. If so, quote them. If not, the mean can be used to compare the order of infusions, but the standard deviation does not make sense. The PCA and Tukey's test don't make sense. I encourage you to indicate the publications that these assessments are quantitative. This will solve a lot of problems.

It seems to me that you should describe the statistical tools used in the study. What they relate to, what they do and in one sentence how they do it. And how to read the results of these analyzes. This is what the chapter Methods is for.

Tukey's test is not described in the text. What is it and what is it for? If you were to describe it, it would turn out that this test belongs to the post-hoc analysis of a much larger test, the Analysis of Variance (ANOVA). Tukey's test decides whether or not the pair means are equal or different if the ANOVA concludes that all means are not the same. Tukey's test cannot be treated as a standalone test. I would do ANOVA for each factor and then do the Tukey test. As long as you can show that the sensory assessments are on a quantitative scale. Otherwise, it is worth using their nonparametric counterparts: the Kruskal-Wallis test and the appropriate post-hoc test (which you can do). And the Tukey test is not LSD but HSD. The LSD test was developed by Fisher. You have to decide which one you use. And by the way: LSD is Least Significant Difference. Not Statistical (line 304).

In my opinion, it is an abuse to use the analysis of variance and Tukey's test to compare the parameters of 6 different infusions. There are 3 pressure values and 2 coffee grade values. These are actually two factors. You should perform a Multivariate Analysis of Variance (MANOVA) with appropriate post-hoc testing and interaction analysis. However, it would be very difficult and impossible to do with three-element samples. So I'm just signaling a problem, but I don't expect you to solve it.

In the chapter "Results and discussion", the results of the study are repeatedly mixed with the theory of the physicochemical phenomena occurring in the infusion of coffee during the formation of coffee cream. It is very interesting, but in my opinion, the description of these phenomena should be included in the earlier part of the article. The results obtained should be discussed here.

The PCA is the largest statistical tool in this work. Greater than ANOVA and Tukey's test combined. The methodology of this tool is not described. Its purpose and interpretation of the results are not either. And the PCA is not easy to interpret. Fortunately, you have clear results. And the interpretation of these results is, in my opinion, very short. Now this test seems to be attached to the article. I know, they asked to add some statisticsJ. Either you have to remove this part (in my opinion, the article will not lose it) or better tell about the results of this analysis.

Detailed and minor remarks:

line 53 - 9 bar? And the study concerns espresso machines with a pressure of 15 or 19 bar. There should probably be "at least 9".

line 210 - 219 - this chapter should be supplemented with descriptions of all statistical tools used in the study. It doesn't have to be in detail, but the basic things need to be described.

line 267 - description of figure 2A. A chart or drawing can show you that something is larger or smaller. The statistical significance of the lower value cannot be stated. You need a test for that. And this drawing doesn't show any test. I would delete the word "statistically".

line 300-305 - the charts are not an analysis. You analyze these charts in the text below. I would remove the word "analysis" from two places in this text.

line 304 - LSD is "significantly" and not "statistically".

line 337-339 - this sentence is quite illegible. Please try to write them down differently.

line 341-342 - what test?

line 342-344 - I do not understand this sentence. How's the statistical analysis? And how was it useful?

line 345-348 - it's true. But why write about it? Think about whether this should be included in the text. As is commonly known, why write about it? And if anything, it should be in the methodology of statistical research. And please do not write that something is common knowledge.

line 366-366 - mean greater than the median. But what follows from this? Is this a description of a right-sided asymmetry? A large number of small bubbles and a small number of large ones? And "wide ranks" can also be told. The values of the described features differ greatly. They are not the same at different pressures. They are varied. This is what the "broad ranks" describe.

line 376 - which test was used to test this statistically significant difference. Where are the results of this test?

line 411 - what test? And what are a, b, c, and d in this table? My guess is the Tukey HSD test groups. But it must be described. And the test is Tukey, not Tuckey (I sometimes make this mistake too, but it needs to be corrected)

Table 4. Do you use standard deviations in any way? If not, I would remove them from the board. It will then become clearer. And there will be no doubt whether standard deviations can be calculated from this data.

line 415-418 - delete the last two sentences. What's on the chart? Arithmetic means?

the line 420 - 19 bar is in the N machine, not the DG.

line 422 - statistically lower values of the profile are shown in Table 1, the Figure shows only the lower values.

Conclusions. It seems to me that one of the most interesting results of the sensory analysis is that coffee brewed at 15 bar for some features scores better than coffee brewed at 19 bar. In the rest, the ratings are the same. So you can say that coffee brewed with 15 bars is better than coffee brewed with 19 bars. It has better organoleptic properties. But the  coffee cream is worse. So it is impossible to say which machine is better, which makes better coffee. And instead of exploiting these results, you switched to different types of Arabica. It is impossible to say anything clever about it, because it is just an example of two types of coffee. It's worth a little work on better conclusions.

I wish you successful work on the article and look forward to the corrected text.

Kind regards

Author Response

REW 1

Dear reviewer, we corrected according your points of observation, as follow:

First: why capsule espresso machines? I can guess the reasons, but it is worth writing about it clearly in the article. The popularity of these coffee machines is probably not a good, or at least the only reason. For example, I do not have a positive opinion of them. Please convince me that it is worth researching them.

Dear Reviewer, according to your request, I have answered as follow (we have inserted the sentences):

Behind the world of coffee, a large, increasingly articulated, and complex business system has developed in Italy and abroad, and its consumption and production are placed between tradition and innovation.

Times have changed for espresso too. The hectic life and the short time, require ever faster and more comfortable systems for preparing coffee, without sacrificing the quality and pleasure that espresso gives.

In recent years, in Italy, the growth of the domestic market for single-portion coffee has been approximately 5.8% more than in previous years, especially with a significant increase in the e-commerce channel, and with greater growth for aluminum capsules compared to paper pods or capsules. The companies engaged in the production of portioned, are in Italy, more than 500. The companies, which have also chosen to include this product in their catalog, work mainly in the production of capsules compatible with the most popular systems. Even if coffee and its preparation methods remain linked to a question of choice due to liking and culture, internationally, in recent years there has been the spread of coffee in capsules and compatible machines. (Coffitalia - Yearbook 2018)

Second: they are different devices. Is pressure the only factor that differentiates it? Isn't there the influence of other factors that differentiate the coffee from these machines? If so, the differences in coffee infusions from these devices cannot be explained by pressure alone. Your research methodology is defensible, but it needs to be written.

Dear Reviewer, the Material and Methods section we explained the characteristics of the devices used for coffee extraction (we included also the quantity of coffee present in the two capsules). We added in the text, as follow:

The coffee capsules, used for Nespresso and Nescafé - Dolce Gusto, had a slightly different shape.

Third: why two types of coffee and why these? If you were to focus on one type of coffee, the pressure effects would be obvious and the results would be clear. Now these are the questions. Why only two species? Are they exhausting all the Arabica varieties? They are so different that the message of the differences is to be visible? And why exactly them? Because the manufacturer has them on offer? A very bad reason. In my opinion, the results shown in Figure 3 show well that this pressure is an ambiguous factor. And changing the type of coffee means that the results are not obvious. Describe it better in your conclusions. It begs for discussion. One species of Arabica is more influenced by pressure than others. Why? What is the conclusion? You have a little bit too few results for a full discussion, but you can signal that you see the problem. And you absolutely have to justify the reason why you took these two types of coffee and just these two. For what?

Dear Reviewer, we added the following sentence:

Coffee production is constantly evolving, in terms of agronomic techniques, selection of new varieties that adapt to different climatic conditions, expressing optimal organoleptic characteristics. Moreover, different varieties of Arabica respond differently to the treatments of the drupes, in post-harvest, showing their characteristics to the best according to the technology applied: "natural coffee" and "washed coffee". The two varieties, discussed here, represent new varietal selections, which correspond to new Arabica coffee crus. The results obtained by the three extraction methods of coffee are reported below.

We added the following sentence:

As reported by Andueza et al, (2007) (3) and Adhikari et al, (2019) (32), even slightly varying the coffee parameters (extraction method, consumption temperature, etc.) there is a quantitative and qualitative variation of volatile and organoleptically active molecules, with consequences on the sensory characteristics of the beverage. In this research, the observed differences are probably due to the distinct extraction of some molecules responsible for the flavor, and the characteristic volatility of some odorous molecules following the three methods applied. All this leads to sensory profiles with distinct shapes (Fig. 3) and statistically significant differences in many of the sensory parameters studied (Table 1).

We added:

Adhikari, J.; Chambers, E.; Koppel, K. Impact of consumption temperature on sensory properties of hot brewed coffee. Food Res. Int. 2019 115 95–104.

Fourth: You used a specific mineral water for the study. Did the mineralization of this water affect the results? Wouldn't a different mineral composition make your results different? Or maybe it has so few minerals that it has no significant effect on the properties of the infusion? It is worth writing.

We added the water composition:

To standardize the tests, we used commercially available water. It presents the following composition: HCO₃ (309 mg/l), Ca⁺⁺ (47,1 mg/l), Mg+ (28,5 mg/l), SiO₂ (17,9 mg/l), NO₃^- (9 mg/l), Na (6,2 mg/l), SO₄^--(4,2 mg/l), Cl^- (2,1 mg/l), K⁺ (1 mg/l); pH 7,5.

We did not test other samples of water, with other water composition, even though the composition of water could affect sensory characteristics of the beverages.

In the study of sensory parameters, a ten-point scale of ratings assigned by panel participants was used. These ratings are not values ​​on a measurable scale! Rating is not a number. Has been given or assigned, not measured. It looks like a number, but it is not. It is a quality feature. They cannot be used to calculate measures and use tools appropriate to quantitative characteristics. Such as means, standard deviations or PCA. I know this is commonly done, but it is incorrect. Do you know of any studies that show that the values ​​of the scale you use are measurable? Or, it is not a mistake to use quantitative tools for these assessments. If so, quote them. If not, the mean can be used to compare the order of infusions, but the standard deviation does not make sense. The PCA and Tukey's test don't make sense. I encourage you to indicate the publications that these assessments are quantitative. This will solve a lot of problems.

We did the QDA in accordance with the article (see chapter "Bibliography):

Andueza, S.; Vila, A. M.; De Peña, M. P.; Cid, C. Influence of coffee/water ratio on the final quality of espresso coffee. J. Sci. Food Agric. 2007, 87, 586–592.

And we measured the intensity ratings (±standard deviations) for sensory attributes.

In “Material and Methods” section I updated the statistical tests applied in this paper. (See below)

I reported the sensory methodology also reported by other authors, not mentioned in this paper, as follow:

1. Susana Andueza, Laura Maeztu, Lucı´a Pascual, Carmen Iba´n˜ez, M Paz de Pen˜a and Concepcio´n Cid. Influence of extraction temperature on the final quality of espresso coffee. J Sci Food Agric 83:240–248 (online: 2003) DOI: 10.1002/jsfa.1304
2. SUSANA ANDUEZA, M. PAZ DE PENA, AND CONCEPCIOÄ N CID. Chemical and Sensorial Characteristics of Espresso Coffee As Affected by Grinding and Torrefacto Roast
3. Agric. Food Chem. 2003, 51, 7034-7039
4. DI DONFRANCESCO, N.GUTIERREZ GUZMAN and E. CHAMBERS IV. COMPARISON OF RESULTS FROM CUPPING AND DESCRIPTIVESENSORY ANALYSIS OF COLOMBIAN BREWED COFFEEB. 2014. Journal of Sensory Studies ISSN 0887-8250

It seems to me that you should describe the statistical tools used in the study. What they relate to, what they do and in one sentence how they do it. And how to read the results of these analyzes. This is what the chapter Methods is for.

Dear Reviewer, we added some information in Material and Methods:

2.5. Statistical analysis

Data were statistically processed by Excel, Statgraphics Centurion XVI (StatPoint Technologies, Inc., 2010) and OriginPro (Graphing & Analysis software, 2022). A descriptive statistical study, using box plots (diameter of foam bubbles) and spider plots (sensory parameters) was conducted; an analysis of variance (ANOVA) was applied for inferential study. Tukey test, HSD (p <0.05) (sensory parameters data) [3], (diameters of foam bubbles) was applied. Principal component analysis (PCA) was applied to identify patterns and structures on the sensory parameters: texture of cream, smell, sour, sweet, bitter, the balance, body, astringent, aroma, and persistence aroma, and the three coffee methods of extraction. For the descriptive analysis, attributes receiving a score <1.0 for all samples were not included in the PCA. To perform the image analysis, about 30 bubbles diameters for each image were measured. All the foaming tests were carried out in triplicate (three individual proofs); in each proof 150 bubbles were considered.

Tukey's test is not described in the text. What is it and what is it for? If you were to describe it, it would turn out that this test belongs to the post-hoc analysis of a much larger test, the Analysis of Variance (ANOVA). Tukey's test decides whether or not the pair means are equal or different if the ANOVA concludes that all means are not the same. Tukey's test cannot be treated as a standalone test. I would do ANOVA for each factor and then do the Tukey test. As long as you can show that the sensory assessments are on a quantitative scale. Otherwise, it is worth using their nonparametric counterparts: the Kruskal-Wallis test and the appropriate post-hoc test (which you can do). And the Tukey test is not LSD but HSD. The LSD test was developed by Fisher. You have to decide which one you use. And by the way: LSD is Least Significant Difference. Not Statistical (line 304).

Dear Reviewer, we corrected row 304. We changed with HSD Tukey test on the caption of Fig. 2.

We added:

Analysis of variance (ANOVA) was applied in all inferential tests.

We followed as reported by

1. Andueza, S.; Vila, A. M.; De Peña, M. P.; Cid, C. Influence of coffee/water ratio on the final quality of espresso coffee. Sci. Food Agric. 2007, 87, 586–592.

In my opinion, it is an abuse to use the analysis of variance and Tukey's test to compare the parameters of 6 different infusions. There are 3 pressure values and 2 coffee grade values. These are actually two factors. You should perform a Multivariate Analysis of Variance (MANOVA) with appropriate post-hoc testing and interaction analysis. However, it would be very difficult and impossible to do with three-element samples. So I'm just signaling a problem, but I don't expect you to solve it.

We followed as reported by

1. Andueza, S.; Vila, A. M.; De Peña, M. P.; Cid, C. Influence of coffee/water ratio on the final quality of espresso coffee. Sci. Food Agric. 2007, 87, 586–592.

In the chapter "Results and discussion", the results of the study are repeatedly mixed with the theory of the physicochemical phenomena occurring in the infusion of coffee during the formation of coffee cream. It is very interesting, but in my opinion, the description of these phenomena should be included in the earlier part of the article. The results obtained should be discussed here.

Dear reviewer, I would leave this part as a deepening of the phenomenon observed here.

The PCA is the largest statistical tool in this work. Greater than ANOVA and Tukey's test combined. The methodology of this tool is not described. Its purpose and interpretation of the results are not either. And the PCA is not easy to interpret. Fortunately, you have clear results. And the interpretation of these results is, in my opinion, very short. Now this test seems to be attached to the article. I know, they asked to add some statisticsJ. Either you have to remove this part (in my opinion, the article will not lose it) or better tell about the results of this analysis.

Dear reviewer, we added the following comments to PCA:

In summary, the PCA (Fig. 4) shows how M Brasile and M Guatemala, extracted with moka, differ from those obtained by capsules at higher pressures (N and DG). With the extraction by moka, it is also possible to distinguish the two varieties (M Brazil, quadrant PC2 positive and M Guatemala, quadrant PC2 negative).

Furthermore, N Guatemala (PC2 positive quadrant) differs from N Brazil (PC2 negative quadrant), highlighting different sensory characteristics of the two varieties of Arabica, with the same extraction method. N Guatemala stands out for its sensory parameters of aroma and balance, while N Brasile stands out for sour, persistence of the aroma, for sweet and for its aromatic intensity.

On the contrary, the two varieties, Guatemala and Brazil, when extracted with the DG method (15 bar) are superimposed, in the PC2 negative quadrant, (between the negative e positive quadrants, PC2); they also share the parameters of texture, sweet, body, astringent and aroma intensity. In this last case, the DG extraction method (15 bar) makes the two coffees very similar to each other.

line 53 - 9 bar? And the study concerns espresso machines with a pressure of 15 or 19 bar. There should probably be "at least 9".

Dear Reviewer, I do not understand this point.

line 210 - 219 - this chapter should be supplemented with descriptions of all statistical tools used in the study. It doesn't have to be in detail, but the basic things need to be described.

This part has already been corrected; the elements you requested have been added.

line 267 - description of figure 2A. A chart or drawing can show you that something is larger or smaller. The statistical significance of the lower value cannot be stated. You need a test for that. And this drawing doesn't show any test. I would delete the word "statistically".

We have delated “statistically”.

line 300-305 - the charts are not an analysis. You analyze these charts in the text below. I would remove the word "analysis" from two places in this text.

We removed “analysis” from the two places.

line 304 - LSD is "significantly" and not "statistically".

Here there was a mistake: it was not LSD but HSD (we have corrected)

line 337-339 - this sentence is quite illegible. Please try to write them down differently.

We changed as follow:

Statistically, the greater is the number of bubbles and the lower is the variability, (many research papers report on very high bubbles number analyzed, also thousands). However, Barik et al. [15], to determinate the minimum number of bubbles, required for the statistical analysis, studied d₃₂ for 50, 100, 150, 200, 250, 300, 350 and 400 bubbles obtained from the optical images.

line 341-342 - what test?

 line 342-344 - I do not understand this sentence. How's the statistical analysis? And how was it useful?

We deleted the sentence:

however, the test carried out in triplicate did not show any difference among the 3 proofs.

line 345-348 - it's true. But why write about it? Think about whether this should be included in the text. As is commonly known, why write about it? And if anything, it should be in the methodology of statistical research. And please do not write that something is common knowledge.

We deleted the sentence:

It is well known that box plots represent the total spread or distribution of data, including minimum and maximum and the outliers. In this research the graphical distribution may provide a way to visualize the range and other characteristics of the analyzed groups of “bubbles sizes” in comparison among them. 

line 366-366 - mean greater than the median. But what follows from this? Is this a description of a right-sided asymmetry? A large number of small bubbles and a small number of large ones? And "wide ranks" can also be told. The values of the described features differ greatly. They are not the same at different pressures. They are varied. This is what the "broad ranks" describe.

This represents disproportion of the diameters of the bubbles and therefore foam instability.

We added:

Therefore, the values of the bubbles diameters change greatly and thus the behavior of the coffee foam differs in function of the applied pressure.

line 411 - what test? And what are a, b, c, and d in this table? My guess is the Tukey HSD test groups. But it must be described. And the test is Tukey, not Tuckey (I sometimes make this mistake too, but it needs to be corrected)

It is Tukey HSD test

We added in the test and on Table 1.

Table 4. Do you use standard deviations in any way? If not, I would remove them from the board. It will then become clearer. And there will be no doubt whether standard deviations can be calculated from this data.

We removed standard deviation on Table 1

line 415-418 - delete the last two sentences. What's on the chart? Arithmetic means?

We deleted the last two sentences.

the line 420 - 19 bar is in the N machine, not the DG.

We corrected

line 422 - statistically lower values of the profile are shown in Table 1, the Figure shows only the lower values.

We removed “statistically”

Conclusions. It seems to me that one of the most interesting results of the sensory analysis is that coffee brewed at 15 bar for some features scores better than coffee brewed at 19 bar. In the rest, the ratings are the same. So you can say that coffee brewed with 15 bars is better than coffee brewed with 19 bars. It has better organoleptic properties. But the  coffee cream is worse. So it is impossible to say which machine is better, which makes better coffee. And instead of exploiting these results, you switched to different types of Arabica. It is impossible to say anything clever about it, because it is just an example of two types of coffee. It's worth a little work on better conclusions.

Dear Reviewer, we added on conclusion section, as follow:

The visual aspect of the coffee cream offers the first information on the sensory and hedonic aspect of the beverage. Espresso coffee in capsules has foam stability which depends on the pressure applied during the extraction of the beverage. High pressures ensure greater stability, and this is also confirmed by the macroscopic and microscopic analysis. In fact, at higher pressures the cream is finer, and the bubbles are smaller, uniform, and stable over time. On the contrary, lower pressures are responsible for disproportionation of the size of the bubbles, and in general for instability.

The sensory analysis of the espresso and moka coffees, seem to confirm how, for the same coffee, the extraction method changes the profile of the beverage's perceptions. Coffee with moka, extracted at low pressures, has much lower sensory profiles than the two espresso coffees obtained at high pressures.

However, the choice of coffee and the assessment of acceptability remains subjective aspects; the texture of the cream, but also its absence as in moka coffee, the release of complex aromas, the bitter taste, and the more or less intense astringency (parameters quantified in this research), represent factors to be analyzed through further investigations by of consumer tests.

Although two theses on coffees in comparison represent a limited number of samples, in this research, by considering the variety of Arabica used, it seems that Guatemala responds differently to the extraction method used; in particular, the finer grain of the bubbles could be responsible for a lower aromatic release when the coffee is extracted at higher pressures. This behavior could be due to the different composition, in lipids, proteins and carbohydrates; and the different composition responds differently according to the applied pressure during the coffee extraction.

In conclusion, coffee is confirmed as an extremely complex beverage, in which even the modification of a parameter can bring different physical and sensory characteristics.

Thank you

Reviewer 2 Report

Article review: Foam characteristics and sensory analysis of Arabica coffee, extracted by espresso capsule and moka methods.

The purpose of this work is to compare three types of coffee extraction: two modes of espresso extraction, with the use of capsules and coffee machines that operate at different pressures, and the method of extraction with the moka. The main purpose of this research is to compare two different coffees, 100% Arabica, produced in Brasile and Guatemala, processed in the same company and extracts with capsules compatible with Nespresso and Dolce Gusto, and with the moka. The crema/foam of the two espresso coffees (moka coffee does not have foam), and the sensory profiles of the coffees, obtained with the three extraction methods, were studied.

The worki is well written. The text clear and easy to read.

The conclusions are consistent with the evidence and arguments presented. The conclusions respond to the set goal of the work.

Author Response

Reviewer 2

There is no observation.

Best Regards

Reviewer 3 Report

The work is relevant from a scientific point of view, although there are many similar works.
I am quite divided in my opinion as to the need for more samples for this work. There is only espresso and moka coffee, and only two origins: Brazil and Guatemala. There are only two pressure variation, 15 and 19 bars. My main doubt is whether this covers a variability that the authors themselves claim exists in the coffee universe. In the conclusion, it is said that "coffee is confirmed as an extremely complex beverage, in which even the modification of a parameter can bring different physical and sensory characteristics". This is more than known, but it is still true. My question is exactly this: so, wasn't a greater range of sample variability necessary? More geographic origins, more variability in pressures, more types of beverage preparation method?
In any case, it is an extremely well-organized, well-written and interesting work.

Author Response

Reviewer 3

Dear reviewer, I agree with your observation: small number of samples, and two different pressures for the extraction of the coffee. However, in accordance with your observations and those of another reviewer, I have specified on the text, the choice of the two varieties; I also stressed the importance of carrying out further investigations on the issues addressed in this paper.

Best Regards

Round 2

Reviewer 1 Report

Dear Authors,
I saw that you made a lot of changes to your article. I hope they will contribute to improving its scientific value.
I have two more minor comments / suggestions. They are not obligatory for you:
In response to the review, you described the sources of the sensory test methodology, and in particular the reasons for using quantitative methods (tests and PCA) to analyze the results. Why didn't you include it in the text? I am convinced, but the readers of the article may not. The recipients of the article are not specialists in sensory analysis. Consider adding a few sentences on this in the body of the article. But it depends on your decision.
Description of the water used in the test. It's probably hotter than it was. I would put the paragraph on water at the end of the subsection because now the descriptions of devices and water are mixed. And the chemical composition of water will tell no one (apart from specialists). What does it mean? Low mineral or high mineral water? Does it affect the quality of the brewed coffee or is it small? The chemical composition itself is unreadable. I think you used this particular water because it has a low mineral content and a neutral taste, which means that this water will not have a significant impact on the resulting infusion. Maybe you want to write about it in one task? But it is also entirely up to you.
Congratulations on your research and description. I wish you further scientific successes.

Author Response

Dear Reviewer

I agree with your observations, but I did not apport any modification to the paper. 

I followed your suggestion to move the paragraph of water composition at the end of the paragraph.

Best Regards